I ask this question because I worry that if an opinion regarding a certain rule holds a majority among users at one point in time, an accepted answer signals that the rule can no longer change to anyone who later considers the question. There is no reason, however, that the rules we put in place must stay till the end of time. Shouldn't we avoid accepting answers to not discourage people from disagreeing with the status quo?

Accepted answers can always be changed, but that doesn't address the point made above. There isn't a clear note on every question reminding potential answerers/voters of this. Additionally, the person who asked the question may become inactive on the site, meaning the accepted answer will never change regardless of any subsequent change in opinion or votes.

  • 4
    \$\begingroup\$ It probably depends on the question type. Some questions ask to make a policy on a topic, I see a reason not to accept an answer there, however other are just asking how the site works, in which case the answers are unlikely to change. \$\endgroup\$ – Post Rock Garf Hunter May 2 '18 at 21:57
  • 2
    \$\begingroup\$ VTC. Should be migrated to meta-meta-PPCG :) \$\endgroup\$ – Khuldraeseth na'Barya May 3 '18 at 15:15

Accepting an answer is a way for the person who asked to question to publicly signal "This is the answer which I found most helpful". Nothing more, nothing less.

So with respect to

an accepted answer signals that the rule can no longer change to anyone who later considers the question

(presumably on a question), the accepted answer never signalled that that was the rule. It may have been on a negative score when it was accepted.

On that basis, I would argue that if there is a reason for not accepting answers on questions it's to avoid confusing people who haven't understood that point. However, without evidence that this is a real problem we don't need a solution.

  • 4
    \$\begingroup\$ I think the biggest disadvantage is that accepting an answer often influences votes by putting the post at the top. How significant that influence is is hard to know and easy to speculate. \$\endgroup\$ – Nathan Merrill May 3 '18 at 14:14
  • \$\begingroup\$ I disagree—rules are often determined by voting behavior on meta. People on main will often appeal to “community consensus” when enforcing rules which simply means this opinion has the most upvotes. One example is whether JavaScript quines can read their functions’ code. It is allowed only because the answer that says they can has more upvotes than the other ones. The recent HTA answer to the rotating quine was deleted by a mod who referenced community consensus as his/her authority to do so. So I think the most upvoted answer in fact often determines a rule. \$\endgroup\$ – dylnan May 3 '18 at 14:52
  • \$\begingroup\$ @dylnan Then? Most upvoted != accepted. \$\endgroup\$ – user202729 May 3 '18 at 14:53
  • \$\begingroup\$ @dylnan, precisely. It's votes (if anything) which matter, not the green tick. \$\endgroup\$ – Peter Taylor May 3 '18 at 14:53
  • 2
    \$\begingroup\$ Well then do we really need a "This is the answer which OP found most helpful" marker? They can upvote just like anyone else. Currently I don't see the accepted marker adding anything but I do see the ways in which it can confuse users \$\endgroup\$ – dylnan May 3 '18 at 15:16
  • \$\begingroup\$ @dylnan Perhaps for the non-policy questions? Also that's an SE ---restriction--- feature, we can't effectively get rid of it (except perhaps by telling everyone on every meta question "don't accept, don't accept", which seems weird) \$\endgroup\$ – user202729 May 4 '18 at 9:52
  • \$\begingroup\$ This is the same as on non-meta. The OP has the right to accept any answer. Although it may be frowned upon by the community, the OP may choose an answer that didn't technically win according to the objective criterion. \$\endgroup\$ – mbomb007 May 10 '18 at 21:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .